UPF algorithm and its application in the GPS/INS integrated navigation

The integration of position and velocity is commonly adopted for observation model in GPS/INS integrated navigation system. But the navigation accuracy of this system is decreased by the correlation between the system´s measurements. So the pseudo-range of the original measurements of GPS was adopted for observation model to eliminate the correlation. To reduce the linearization error, the unscented particle filter (UPF) was introduced to estimate the system states. The simulation was carried out under the condition of less than four GPS satellites data. The simulation results show that the estimated values with UPF method are quite close to the real values and the estimation error is much smaller than the inherent error of GPS. Both the figures and the statistical results prove that the UPF method present high navigation accuracy when employed in GPS/INS integrated navigation system.